Implementing a Smart Beta Index: The Implications of a Dual Performance Objective and Limited Liquidity

Abstract

A smart beta portfolio that tracks an explicit smart beta index has two performance benchmarks: the smart beta index itself and a cap-weighted index that represents the broader equity market. This dual performance objective creates a trade-off whenever the portfolio is large relative to the liquidity of the smart beta index. A portfolio that fully replicates the smart beta index, and trades all shares near the close on each rebalancing date of the smart beta index, can incur significant transaction costs and thereby adversely affect the performance of the smart beta index vis-à-vis the cap-weighted market. This article shows that sampling (by excluding the least-liquid stocks) and trading patiently (by trading over a longer interval of time) can reduce expected transaction costs and improve the performance of the smart beta index (and, by extension, the smart beta portfolio). Sampling and patient trading cause tracking error between the smart beta portfolio and the smart beta index. However, this tracking error, when modest, has very little impact on the tracking error between the smart beta portfolio and the cap-weighted performance index. The authors conclude that the optimal level of tracking error between the smart beta portfolio and the smart beta index may be much higher than the level usually chosen for a cap-weighted passive portfolio.